Headless commerce development: a custom storefront on your existing engine
Headless commerce keeps your commerce engine (products, stock, cart, checkout, orders) and replaces the storefront with separate software built to your design. WavX builds headless storefronts in Next.js or Shopify's Hydrogen. Most stores do not need one: a well-built theme is cheaper to build and to run, and this page sets out how to tell the difference.
What headless changes, and what stays
A headless store is two systems. The commerce engine, usually Shopify or a custom back end, keeps doing what it does well: the catalogue, stock, prices, cart, checkout, payments and orders. The storefront, the pages customers see, becomes a separate application that reads from the engine through its API and draws every page itself.
| Stays in the engine | Moves to the storefront you own |
|---|---|
| Products, variants, stock, prices | Page design and layout, without a theme's limits |
| Checkout and payment | How pages load, what is cached, what scripts run |
| Orders, refunds, customers in the admin | Content structure: lookbooks, guides, comparison pages |
| Discounts and tax rules | Search and filtering, if the built-in ones are not enough |
| Back-office apps that work through the admin | Any widget that used to come from a theme app |
On Shopify, the storefront reads from the Storefront API and hands the cart to Shopify's checkout through a checkout URL. Shopify's documentation, read on 2 October 2026, says requests from real buyers are not held to a fixed per-minute limit, while checkout creation is throttled. Tokens come in two kinds: public tokens used from a browser or app, and private tokens that must stay on a server.
When headless is worth it
Headless is a trade: more freedom over the storefront, in exchange for a second system to build, host and maintain. It is worth it in a few specific situations.
| Situation | Headless? |
|---|---|
| A standard store that a good theme serves well | No. Improve the theme |
| The store is slow and the cause is apps and scripts loaded by the theme | Usually no. Remove or replace the apps first; see Shopify speed optimisation |
| Content-led brand: long guides, editorial pages, rich product stories that a theme fights | Often yes |
| One engine feeding a website, a mobile app and other channels | Yes |
| Product data comes from several systems (Shopify plus a PIM, ERP or custom pricing service) | Yes |
| Several country or brand sites from one catalogue | Often yes |
| A small team with no developer after launch | No |
Shopify's own description of headless is "all the power of Shopify under the hood, with complete control over the frontend". That control is what you pay for. If you would not use it, do not buy it. When headless Shopify is worth it works through the decision in more depth.
Hydrogen, Next.js or another framework
| Option | Fits when | Notes |
|---|---|---|
| Hydrogen on Oxygen | The engine is Shopify and you want Shopify's own components and hosting | Shopify states Hydrogen is built on React Router, and that Oxygen hosting is available at no extra charge on its paid plans (Starter, Basic, Grow, Advanced, Plus and Pause and build) but not on Agentic plans, as of October 2026 |
| Next.js on Vercel, AWS or Google Cloud | The store also draws on other systems, the back end is not Shopify, or the rest of your web estate is in Next.js | Hosting is billed by the provider you choose |
| Custom back end plus custom storefront | Your pricing, catalogue or ordering rules do not fit a platform at all | This is a custom store, not headless on a platform; see e-commerce development |
Shopify's documentation also says any framework, tooling or hosting can be used with its APIs. The choice is mostly about hosting, your team and the systems around the store, not about the framework itself.
What WavX builds in a headless project
| Workstream | What it covers |
|---|---|
| Audit of the current store | Speed measured on real devices, the app list, which apps touch the storefront, and what each would need in a headless build |
| Storefront | Home, collection, product, search, cart, account and content pages built in Next.js or Hydrogen |
| Content management | A content system for marketing pages, so the team is not dependent on a developer for every banner |
| App replacements | Reviews, wishlists, upsells, size guides and similar, either rebuilt or connected through their APIs |
| Checkout hand-off | Cart to the platform's checkout, with discounts, gift cards and analytics carried across |
| Analytics and tracking | GA4 and ad pixels re-implemented, because the theme's built-in tracking no longer applies; see GA4 integration |
| SEO carry-over | URL structure kept or redirected, metadata, structured data and sitemaps re-created |
| Performance | Measured against Core Web Vitals before launch, not after |
The SEO carry-over deserves its own line in any proposal. A headless rebuild changes how every page is generated. If URLs, canonical tags or structured data are lost on the way, search traffic can drop after launch. The checks are the same as for a site migration.
What it costs: a planning range
These figures come from the WavX software cost estimator, treating the storefront as a custom web application connected to one commerce engine. They are planning ranges, not quotes.
| Configuration | Cost model inputs | Subtotal | Planning range | Estimator timeline |
|---|---|---|---|---|
| Storefront on an existing engine, standard design | Custom web application ₹4,50,000 + third-party integrations ₹60,000 | ₹5,10,000 | ₹4,33,500 to ₹6,37,500 | 8–14 weeks |
| As above, with customer accounts and custom design | (₹4,50,000 + integrations ₹60,000 + auth and roles ₹60,000) × 1.25 custom design | ₹7,12,500 | ₹6,05,625 to ₹8,90,625 | 8–14 weeks |
| As above, plus a companion mobile app on the same engine | (₹5,70,000 + companion mobile app ₹2,00,000) × 1.25 custom design | ₹9,62,500 | ₹8,18,125 to ₹12,03,125 | 14–22 weeks |
Assumptions: small scale (up to about 1,000 users) in the estimator's terms; it multiplies by 1.3 for a growing user base. The figure does not include rebuilding a long list of theme apps, which is quoted after the audit, or the platform's own subscription.
Running costs change too. You pay for storefront hosting (unless it is Oxygen), for the content system if it is a paid one, and for maintenance on a second codebase. WavX's published maintenance band is 15 to 25% of the build cost per year; see app maintenance cost per year.
How the work runs
- Audit. Measure the current store, list every app and what it does on the storefront, and write down why headless is being considered. The audit can end with "fix the theme".
- Architecture. Choose the framework, hosting and content system; map every page type and URL.
- Build in parallel. The headless storefront runs on a staging address while the live store keeps trading.
- Parity check. Every product, discount type, tracking event and redirect tested against the live store.
- Cut-over. Domain moved to the new storefront at a quiet time, with the old theme kept ready to switch back.
- Watch. Search Console, analytics and conversion data compared with the weeks before launch.
When not to go headless
- Your store is under a few hundred products and a theme does what you need. A custom theme costs less and your team keeps the theme editor.
- The goal is speed alone. Most slow Shopify stores are slow because of apps and images, which can be fixed without a rebuild. See why a Shopify store is slow.
- No one will maintain it. A headless storefront is software. If there is no developer or maintenance arrangement after launch, it will fall behind the platform.
- Trading depends on many theme apps. Each one becomes a small integration project. Count them before deciding.
Headless builds are part of WavX's e-commerce development work, alongside custom storefronts and B2B portals.
Frequently asked questions
Does going headless make a store faster?
It can, because the storefront is built and tuned for one store. It does not by default. A headless storefront that loads large scripts or fetches data badly is slower than a good theme. We measure the current store before recommending a rebuild, and often the answer is to fix the theme.
Do we keep the Shopify admin and checkout?
Yes. Products, inventory, orders and checkout stay in Shopify, and your team uses the same admin. The storefront hands the cart to Shopify's checkout. What changes is that design and content changes on the storefront need a developer or a content system, not the theme editor.
Hydrogen or Next.js?
Hydrogen if the engine is Shopify and you want Shopify's own components and hosting, which Shopify says is included at no extra charge on paid plans. Next.js if the store also draws on other systems, if the back end is not Shopify, or if your team already works in it. Both are React; the choice is about hosting and the rest of your stack.
How much does a headless storefront cost?
From the WavX software cost estimator, a storefront on an existing commerce engine works out to a planning range of about ₹4.3 lakh to ₹6.4 lakh at standard design, and about ₹6.1 lakh to ₹8.9 lakh with customer accounts and custom design. These are planning ranges, not quotes. Running costs come on top.
Will our existing Shopify apps keep working?
Apps that work in the admin or the back end usually do. Apps that inject something into the theme, such as reviews widgets, upsell blocks or chat, do not appear on a headless storefront by themselves. Each one has to be rebuilt or connected through its API, and that list is a large part of the scope.
Sources
- Shopify developer documentation: Headless storefronts · read 2 October 2026
- Shopify developer documentation: Hydrogen fundamentals (Oxygen availability) · read 2 October 2026
- Shopify developer documentation: Storefront API · read 2 October 2026
Related
Build your own software — your way, your pricing.
WavX Solutions is here to create your own software in a fully custom way, built exactly how you work — with a pricing model that fits your business. Connect now and let's build it.